Data-driven Irrigation Water Resource Scheduling Method and System

Abstract

This invention discloses a data-based prediction-based method and system for irrigation water resource scheduling, belonging to the field of resource scheduling and intelligent decision-making. The method includes: acquiring historical irrigation water consumption, historical temperature data, weather forecast data, and real-time water source status data; using exponential smoothing to recursively deduce the baseline water demand, combined with temperature regression correction to obtain the predicted irrigation water demand; establishing triple water source constraints: single-period, cumulative, and dynamic changes; constructing a multi-objective optimization model by weighted summation of deficit, waste, and water supply costs after benchmarking; using a genetic algorithm to solve for the decision water supply, and generating a scheduling scheme based on the hourly water consumption coefficient, followed by rolling optimization. This invention achieves a unified approach to demand forecasting, constrained feasible scheduling, and multi-objective global optimization.

Description

Technical Field

[0001] This invention belongs to the field of resource scheduling and intelligent decision-making, specifically relating to a method and system for scheduling irrigation water resources based on data prediction. Background Technology

[0002] With the in-depth development of smart city and refined resource management technologies, the scheduling and management of irrigation water resources in cities and parks are facing increasingly complex challenges. In scenarios such as urban greening, park landscape maintenance, and municipal public area maintenance, multiple irrigation areas share a single water supply system. The irrigation water demand of each area varies in time and space and is affected by factors such as season, weather, and human activity. Scientific decision-making is required to achieve efficient allocation of irrigation water resources.

[0003] Existing irrigation water resource scheduling technologies mainly employ the following methods: fixed quota scheduling; threshold-triggered scheduling; and manual experience-based scheduling. These technologies share several common drawbacks: First, they lack predictive capabilities, failing to forecast water demand in various irrigation areas over a future period based on historical data, meteorological information, and regional activities, leading to delayed scheduling decisions. Second, they neglect water source constraints, failing to consider the real-time water storage and replenishment capacity of irrigation sources, potentially resulting in on-demand water supply even when water resources are insufficient, causing water depletion or supply interruptions. Third, they lack sufficient multi-objective consideration, failing to comprehensively balance multiple objectives such as irrigation demand satisfaction rate, water resource utilization efficiency, and water supply costs, making it difficult to achieve globally optimal scheduling. Therefore, a technical solution is urgently needed that can predict water demand in various irrigation areas based on multi-source data and achieve dynamic optimization scheduling of irrigation water resources under water source constraints. Summary of the Invention

[0004] The purpose of this invention is to solve the technical problems of lagging scheduling decisions, neglecting water source constraints, and insufficient multi-objective trade-offs in the prior art.

[0005] To achieve the above objectives, the present invention employs the following technical solution:

[0006] In a first aspect, the present invention provides a method for irrigation water resource scheduling based on data prediction, the method comprising the following steps:

[0007] Step S1: Obtain the hourly historical irrigation water consumption and historical temperature data for each irrigation area for 90 consecutive days before the current moment, obtain the weather forecast data for 7 consecutive days after the current moment, and obtain the real-time water storage and replenishment rate of the irrigation water source.

[0008] Step S2: Based on the hourly historical irrigation water consumption, the daily historical irrigation water consumption is obtained by summarizing the daily data. The baseline irrigation water demand for each irrigation area is calculated recursively using the exponential smoothing method. A temperature influence correction model is established based on the historical temperature data, and the temperature influence increment is calculated by combining the forecast temperature in the meteorological forecast data. The baseline irrigation water demand and the temperature influence increment are superimposed to obtain the predicted irrigation water demand for each irrigation area.

[0009] Step S3: Based on the real-time water storage, replenishment rate and forecast precipitation in the meteorological forecast data, calculate the daily available water volume, establish a single-period water source constraint based on the daily available water volume, establish a multi-period cumulative water source constraint based on the current water storage and cumulative replenishment, and establish a dynamic change equation for water storage that describes the daily change pattern of water source water storage.

[0010] Step S4: Calculate the irrigation demand deficit and water waste based on the predicted irrigation water demand and the decided water supply; calculate the water supply cost based on the decided water supply and the unit water supply cost; and construct a benchmarked multi-objective optimization scheduling model by performing benchmarked dimensionless processing on the irrigation demand deficit, water waste, and water supply cost respectively.

[0011] Step S5: Using the single-period water source constraint, multi-period cumulative water source constraint, water storage dynamic change equation, water supply non-negativity constraint, and minimum water supply constraint as complete constraints, the genetic algorithm is used to solve the benchmark multi-objective optimization scheduling model to obtain the decision water supply and water supply priority for each irrigation area.

[0012] Step S6: Based on the decision water supply volume and water supply priority, and combined with the hourly water consumption coefficient obtained from the hourly historical irrigation water consumption statistics, allocate the daily decision water supply volume to each hour to generate a water supply schedule and scheduling decision scheme and output it; when the preset update cycle is reached, return to step S1 for rolling optimization.

[0013] Furthermore, the hourly historical irrigation water consumption refers to the hourly irrigation water consumption of each irrigation area for the consecutive 90 days prior to the current moment, collected hourly by smart water meters installed on the water supply pipelines of each irrigation area; the historical temperature data refers to the daily average temperature for the consecutive 90 days prior to the current moment; the meteorological forecast data includes the forecast daily maximum temperature, forecast daily minimum temperature, and forecast precipitation for the consecutive 7 days following the current moment; the real-time water storage is read by a level sensor installed in the irrigation water source, and the replenishment rate is read by a flow meter installed on the replenishment pipeline.

[0014] Furthermore, the recursive calculation method for the baseline irrigation water requirement is as follows:

[0015] The hourly historical irrigation water consumption is summarized on a daily basis to obtain the daily historical irrigation water consumption;

[0016] ;

[0017] in For the first The irrigation area The standard irrigation water requirement per day, in cubic meters. 3 , For the first The irrigation area Daily historical irrigation water consumption per day, in cubic meters. 3 , The smoothing coefficient is dimensionless. For the first The irrigation area The baseline irrigation water requirement for the day; when hour, Take the average of the historical daily irrigation water usage over the seven consecutive days preceding the current moment. The value ranges from 0.2 to 0.4, and the values ​​are iterated in increments of 0.01. The optimal value is determined by minimizing the fitting error of the historical data for the 90 consecutive days prior to the current moment.

[0018] Furthermore, the method for establishing the temperature influence correction model is as follows:

[0019] For the A least-squares linear regression was performed on the historical daily average temperature of the 30 consecutive days preceding the current time in an irrigation area and the historical daily irrigation water consumption of that area to obtain the temperature influence coefficient and the linear regression intercept; the historical baseline temperature was calculated as the arithmetic mean of the historical daily average temperatures of the 30 consecutive days; the baseline correction was obtained by converting the regression intercept and the temperature influence coefficient; The future of the region The method for calculating the incremental impact of daytime temperature is as follows:

[0020] ;

[0021] in, For the first The future of the region The temperature of the day affects the increment, unit: m 3 , For the future Forecast daily average temperature for the day, unit: °C. Historical baseline temperature, unit: °C Temperature influence coefficient, unit: m 3 / ℃, This is the baseline correction amount, in meters. 3 ,Depend on The conversion yields, This is the intercept for linear regression.

[0022] Furthermore, the method for calculating the predicted irrigation water demand is as follows:

[0023] ;

[0024] in For the first The future of the region Predicted irrigation water demand per day, unit: m³ 3 , Ensure that the predicted value is non-negative. For the first The future of the region The temperature of the day affects the increment, unit: m 3 , For the first The irrigation area The standard irrigation water requirement per day, in cubic meters. 3 .

[0025] Furthermore, the calculation of the daily available water volume is as follows:

[0026] ;

[0027] in, For the first Daily available water volume For the first Daily water storage capacity, unit: m³ 3 , Replenishment rate, unit: m 3 / sky; The time step is in days, and we take 1 day. For the first Rainfall replenishment, unit: m 3 It is calculated by combining the forecast precipitation with the catchment area and runoff coefficient;

[0028] The single-time water source constraint is:

[0029] ;

[0030] in, For the first The first region Daily water supply volume, unit: m³ 3 , are decision variables in the optimization model; This represents the total number of irrigated areas.

[0031] The multi-time period cumulative water source constraint is:

[0032] ;

[0033] in, The water volume on the first day is recorded as the water volume at the current moment.

[0034] The equation for the dynamic change of water storage is:

[0035] ;

[0036] in, Indicates the first Daily water storage capacity, unit: m³ 3 , To optimize the total number of time periods, .

[0037] Furthermore, the irrigation demand deficit is:

[0038] ;

[0039] The amount of water wasted is:

[0040] ;

[0041] The water supply cost is:

[0042] ;

[0043] The objective function of the benchmark multi-objective optimization scheduling model is:

[0044] ;

[0045] in, This represents the irrigation deficit, expressed in meters. 3 , For the first The first region Predicted irrigation water demand per day, in cubic meters. 3 ; For the first The first region Daily water supply volume, in cubic meters 3 , This represents the total number of irrigated areas. To optimize the total number of time periods, ; Water wastage, expressed in cubic meters (m³) 3 , For water supply costs, For the first The first region Daily unit water supply cost, unit: yuan / m³ 3 It is calculated by adding energy consumption costs to the local water fee standard; The overall target value is dimensionless. These are the first, second, and third weighting coefficients, which are dimensionless and determined using the analytic hierarchy process (AHP), satisfying the following conditions: , This serves as a baseline value for irrigation demand deficit. This serves as a benchmark for water waste. This is the benchmark value for water supply costs; when , or When the value is 0, the corresponding baseline value is taken as 1.

[0046] Furthermore, the method for determining the benchmark value is as follows:

[0047] The baseline value for irrigation demand deficit is the sum of the predicted irrigation water demand for each region over seven consecutive days.

[0048] The baseline value for water waste is the sum of the products of the maximum daily historical irrigation water consumption in each region over 90 consecutive days and 7.

[0049] The benchmark value for water supply cost is the sum of the products of the maximum daily historical irrigation water consumption in each region over 90 consecutive days and the unit water supply cost.

[0050] Furthermore, the complete constraints also include a non-negative water supply constraint and a minimum water supply constraint, wherein the minimum water supply is taken as 70% to 80% of the lowest daily historical irrigation water consumption over the previous 90 consecutive days; the parameters of the genetic algorithm are: population size is 5 to 10 times the dimension of the decision variable, crossover probability is 0.7 to 0.9, mutation probability is 0.05 to 0.15, and the maximum number of iterations is 800 to 1500 generations;

[0051] The hourly water supply allocation formula is as follows:

[0052] ;

[0053] in For the first The first region Heavenly Hourly water supply allocation, unit: m 3 , For the first The first region The hourly water consumption coefficient is dimensionless and is obtained from the statistical data of historical irrigation water consumption over 90 consecutive hours.

[0054] The preset update cycle is 1 day, which is determined based on the update frequency of meteorological forecast data and the engineering practice of using daily as the basic decision-making unit for irrigation scheduling.

[0055] A second aspect of the present invention provides an irrigation water resource scheduling system based on data prediction, the system comprising the following modules: a data acquisition module, a water demand prediction module, a water source constraint module, a model building module, an optimization solution module, and a scheme output module, wherein:

[0056] The data acquisition module is used to obtain hourly historical irrigation water consumption and historical temperature data for each irrigation area for 90 consecutive days before the current moment, obtain meteorological forecast data for 7 consecutive days after the current moment, and obtain real-time water storage and replenishment rate of irrigation water sources.

[0057] The water demand forecasting module is used to obtain the daily historical irrigation water volume by summarizing the hourly historical irrigation water volume on a daily basis, and to recursively calculate the baseline irrigation water demand for each irrigation area using the exponential smoothing method; to establish a temperature impact correction model based on the historical temperature data, and to calculate the temperature impact increment by combining the forecast temperature in the meteorological forecast data; and to superimpose the baseline irrigation water demand with the temperature impact increment to obtain the predicted irrigation water demand for each irrigation area.

[0058] The water source constraint module is used to calculate the daily available water volume based on the real-time water storage, replenishment rate and forecast precipitation in the meteorological forecast data, establish a single-period water source constraint based on the daily available water volume, establish a multi-period cumulative water source constraint based on the current water storage and cumulative replenishment, and establish a dynamic change equation of water storage describing the daily change law of water source water storage.

[0059] The model building module is used to calculate the irrigation demand deficit and water waste based on the predicted irrigation water demand and the decision variables, taking the water supply volume of each irrigation area at each time period as the decision variables to be optimized; calculate the water supply cost based on the decision variables and the unit water supply cost; and construct a benchmarked multi-objective optimization scheduling model by performing benchmarked dimensionless processing on the irrigation demand deficit, water waste and water supply cost respectively.

[0060] The optimization solution module is used to solve the benchmarked multi-objective optimization scheduling model using a genetic algorithm, taking the single-time water source constraint, multi-time cumulative water source constraint, water storage dynamic change equation, water supply non-negativity constraint, and minimum water supply constraint as complete constraints, to obtain the decision water supply and water supply priority for each irrigation area.

[0061] The scheme output module is used to allocate the daily decision water supply to each hour based on the decision water supply volume and water supply priority, combined with the hourly water consumption coefficient obtained from the hourly historical irrigation water consumption statistics, to generate a water supply schedule and scheduling decision scheme and output it; when the preset update cycle is reached, it returns to the data acquisition module for rolling optimization.

[0062] Compared with the prior art, the present invention has the following beneficial effects:

[0063] First, this invention achieves forward-looking and accurate prediction of irrigation water demand. It utilizes historical irrigation water consumption and historical temperature data, employing exponential smoothing and least squares linear regression to establish a baseline irrigation water demand recursive model and a temperature-corrected model, respectively. Combined with forecasted temperatures from meteorological data, it calculates the predicted irrigation water demand for each irrigation area over the next 7 days. Specifically, the smoothing coefficient of the exponential smoothing method is automatically optimized within the range of 0.2 to 0.4 using a grid search method, aiming to minimize the fitting error of historical data. The temperature regression model is dynamically updated based on rolling 30-day historical data, enabling the prediction model to adaptively track changes in water consumption patterns in each irrigation area. Unlike existing passive scheduling methods that rely on fixed quotas or threshold triggers, this invention achieves forward-looking quantitative prediction of future water demand, transforming scheduling decisions from reactive post-event responses to proactive pre-event planning.

[0064] Secondly, a complete dynamic water source constraint system was constructed to ensure the feasibility of the scheduling plan. Based on real-time water storage, replenishment rate, and forecasted precipitation, this invention establishes a triple constraint mechanism: single-period water source constraint, multi-period cumulative water source constraint, and a dynamic water storage change equation. The single-period water source constraint ensures that the daily water supply does not exceed the available water volume for that day. The multi-period cumulative water source constraint ensures that the cumulative water supply over the entire optimization cycle does not exceed the sum of the current water storage and the cumulative replenishment. The dynamic water storage change equation describes the daily evolution of water storage. By organically combining these three constraints, the scheduling plan is ensured to remain within the water supply capacity range from three dimensions: daily water volume limit, full-cycle cumulative water use limit, and dynamic water volume evolution. This effectively avoids the problem of water source depletion or water supply interruption due to excessive water use in the early stages. Simultaneously, this invention adds a minimum water supply constraint for each irrigation area, taking 70%–80% of the historical lowest water use as the minimum guaranteed water supply, forcibly retaining the basic irrigation water required for vegetation survival, while also considering the ecological safety baseline.

[0065] Third, it achieves multi-objective comprehensive trade-offs and global optimal scheduling. This invention simultaneously incorporates irrigation demand deficit, water resource waste, and water supply cost into the optimization objectives. It employs a dimensionless processing method based on extreme scenarios, using the maximum deficit under no water supply, the historical maximum water supply, and the corresponding maximum water supply cost as benchmark values. This unifies the three sub-objectives with different dimensions into dimensionless scalars, solving the problem in existing technologies where demand satisfaction rate, water resource utilization rate, and water supply economics cannot be directly compared and comprehensively balanced. Based on this, the analytic hierarchy process (AHP) is used to determine the weight coefficients of each sub-objective. Through pairwise comparisons of the judgment matrix by experts and consistency checks, the weight allocation has a theoretical basis and interpretability, avoiding the subjectivity of arbitrary weight assignment and achieving globally optimal scheduling of three conflicting objectives within the same framework.

[0066] Fourth, this invention generates an hourly-level refined scheduling scheme that can be directly implemented and establishes a rolling optimization mechanism. Based on the daily water supply decision obtained from a genetic algorithm and the water supply priority of each region, combined with the hourly water consumption coefficient obtained from hourly historical irrigation water consumption statistics, the daily water supply is allocated to each hour. This generates a complete scheduling decision scheme that includes a water supply schedule, time-segmented water supply suggestions, water supply priorities, and water source status early warning information, allowing the optimization results to be directly converted into executable operational instructions for engineering projects. Simultaneously, this invention sets a preset update cycle. When the cycle arrives, the latest data is automatically collected and the entire optimization process is re-executed. This enables the scheduling scheme to be dynamically adjusted based on weather forecast updates and actual water consumption feedback, maintaining the continuous optimality of the decision and improving the system's engineering practicality and environmental adaptability. Attached Figure Description

[0067] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the embodiments will be briefly introduced below. It should be understood that the following drawings only show some embodiments of the present invention and should not be regarded as a limitation on the scope. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.

[0068] Figure 1 This is a schematic diagram of the data prediction-based irrigation water resource scheduling method of the present invention;

[0069] Figure 2 This is a schematic diagram of the structure of the irrigation water resource scheduling system based on data prediction according to the present invention. Detailed Implementation

[0070] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. The components of the embodiments of the present invention described and shown in the accompanying drawings can generally be arranged and designed in various different configurations.

[0071] Therefore, the following detailed description of the embodiments of the invention provided in the accompanying drawings is not intended to limit the scope of the claimed invention, but merely to illustrate selected embodiments of the invention. All other embodiments obtained by those skilled in the art based on the embodiments of the invention without inventive effort are within the scope of protection of the invention.

[0072] It should be noted that similar labels and letters in the following figures indicate similar items. Therefore, once an item is defined in one figure, it does not need to be further defined and explained in subsequent figures.

[0073] In the description of the embodiments of the present invention, it should be noted that if terms such as "upper," "lower," "horizontal," or "inner" indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings, or the orientation or positional relationship commonly used when the product of the invention is in use, they are only for the convenience of describing the present invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation, and therefore should not be construed as a limitation of the present invention. Furthermore, terms such as "first" and "second" are only used to distinguish descriptions and should not be construed as indicating or implying relative importance.

[0074] Furthermore, the use of the term "horizontal" does not imply that the component must be absolutely horizontal, but rather that it can be slightly tilted. For example, "horizontal" simply means that its direction is more horizontal than "vertical," and does not mean that the structure must be completely horizontal, but can be slightly tilted.

[0075] In the description of the embodiments of the present invention, it should also be noted that, unless otherwise explicitly specified and limited, the terms "set," "install," "connect," and "link" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; and they can refer to the internal connection of two components. Those skilled in the art can understand the specific meaning of the above terms in the present invention according to the specific circumstances.

[0076] The present invention will now be described in further detail with reference to the accompanying drawings:

[0077] See Figure 1 This invention discloses a data-based prediction-based irrigation water resource scheduling method, comprising:

[0078] Step S1: Obtain historical irrigation water consumption data for 90 consecutive days, historical temperature data for 90 consecutive days, weather forecast data for 7 consecutive days, and real-time status data of irrigation water sources.

[0079] It should be noted that irrigation water demand is influenced by historical water usage patterns and temperature changes, and is also constrained by the availability of water resources. This step obtains all the necessary data: historical irrigation water usage and historical temperature data are used for modeling and extrapolation, meteorological forecast data are used to predict future water demand, and real-time water storage and replenishment rates are used to establish water resource constraints.

[0080] Collect hourly historical irrigation water volume for the 90 consecutive days prior to the current moment. Smart water meters are installed in the water supply pipelines of each irrigation area to continuously collect hourly irrigation water consumption data for 90 consecutive days, with the current time as the baseline. The unit is m³. 3 ,in For area code, For date numbering, This indicates the 90th day prior to the current time. It means yesterday, Hour numbering ( The daily historical irrigation water consumption is summarized on a daily basis. The system collects the historical daily average temperature for the 90 consecutive days prior to the current moment, in °C; and the system collects meteorological forecast data for the 7 consecutive days following the current moment, including daily maximum temperature ( °C), daily minimum temperature ( °C), and precipitation ( mm), obtained via meteorological API. The forecast daily average temperature is obtained by averaging the daily maximum and daily minimum temperatures. ( ); collect real-time water storage of irrigation water sources Unit: m 3 The water level is read by a level sensor installed in the irrigation water source, and the replenishment rate is read by a flow meter installed on the replenishment pipeline. Unit: m 3 / day; the replenishment rate remains constant within the optimization cycle, i.e. ( If the replenishment rate exhibits a known daily variation pattern, such as differing between weekdays and weekends, then the value is taken daily according to the actual pattern.

[0081] In one embodiment of the present invention, taking urban park greening irrigation as an example, it includes a lawn area. Garden area Forest area The current time is 00:00 on August 1, 2025. Smart water meters will be installed in the water supply pipelines of various areas to collect hourly historical irrigation water consumption data from May 3 to August 1, 2025. It also records the historical daily temperatures for the same period; it obtains meteorological forecast data from August 2nd to August 8th, 2025 through the meteorological API, with a forecast daily high of 34℃, a forecast daily low of 30℃, and a forecast precipitation of 0mm; and it reads the current real-time water storage volume through the level gauge. m 3 Read the current supply rate through the flow meter. m 3 / day, and this rate remains constant during the 7-day optimization period.

[0082] Step S2: The baseline irrigation water demand is calculated recursively using the exponential smoothing method. The temperature influence correction model is established using least squares linear regression and the temperature influence increment is calculated. The predicted irrigation water demand is obtained by superimposing the results.

[0083] Existing technologies typically employ reactive scheduling decisions after water shortages occur, lacking forward-looking assessments of future water demand. This step recursively extrapolates baseline water usage levels based on historical water use patterns, establishes a temperature correction model based on the regression relationship between historical temperature and water consumption, calculates the incremental impact of temperature using weather forecasts, and comprehensively obtains the predicted irrigation water demand for each region over the next 7 days. This step processes historical water consumption and temperature data through two parallel sub-steps, S2.1 and S2.2, respectively, and finally outputs the combined data in S2.3.

[0084] Step S2.1: Calculate the baseline irrigation water demand recursively using the exponential smoothing method, and statistically determine the hourly water consumption coefficient based on the hourly historical irrigation water consumption over 90 consecutive days. The specific method is as follows:

[0085] ;

[0086] in For the first The irrigation area The standard irrigation water requirement per day, in cubic meters. 3 , For the first The irrigation area Daily historical irrigation water consumption per day, in cubic meters. 3 , The smoothing coefficient is dimensionless. For the first The irrigation area The baseline irrigation water requirement for the day; when hour, Take the average daily historical irrigation water consumption over the seven consecutive days preceding the current moment.

[0087] It should be noted that exponential smoothing assigns a higher weight to recent data by weighting historical data, with the weight being 1 / 2. It can effectively filter out random noise and extract a relatively stable baseline water level. The larger the value, the faster the model responds to recent changes; The smaller the value, the smoother the predicted value. Initial baseline value. Take the average daily water consumption of the 7 consecutive days preceding the current moment to give the recursive formula an initial value.

[0088] It should be further explained that, The value range is 0.2 to 0.4, derived from engineering practice experience with exponential smoothing. Those skilled in the art use a grid search method to iterate through the range of 0.2 to 0.4 with a step size of 0.01, fitting the error using historical data from the previous 90 consecutive days. Minimize to determine the optimal value for the objective.

[0089] Based on the hourly historical irrigation water volume over the previous 90 consecutive days. Statistical hourly water consumption coefficient ; ;

[0090] This coefficient is a dimensionless proportional value that reflects the distribution pattern of historical actual water consumption throughout the day.

[0091] In one embodiment of the present invention: the current time is 00:00 on August 1, 2025. . Take the average daily water consumption of 115m³ for the seven consecutive days prior to August 1st (July 26th to August 1st). 3 Historical daily irrigation water usage in Region A yesterday (August 1st). m 3 The baseline irrigation water demand for the first day (August 2nd) is as follows. m 3 .

[0092] Step S2.2: Perform least squares linear regression on the historical temperature and water consumption over 30 consecutive days, and calculate the incremental temperature impact by combining it with the forecast temperature.

[0093] Based on the historical daily average temperature over the 30 consecutive days preceding the current moment The temperature influence coefficient was obtained by performing a least-squares linear regression with the historical daily irrigation water volume. and benchmark correction amount The method for calculating the temperature impact increment for each day of the next 7 days, combined with the forecast daily average temperature, is as follows:

[0094] By analyzing the historical daily average temperature over the 30 consecutive days preceding the current moment. ( (and historical daily irrigation water consumption for each region) Perform least squares linear regression:

[0095] ;

[0096] in, For the first The first region Historical irrigation water usage Temperature influence coefficient, unit: m 3 / ℃, This is the linear regression intercept, in meters. 3 ;

[0097] Historical benchmark temperature , Unit: ℃.

[0098] Depend on The baseline correction amount is obtained by conversion. Unit: m 3 , so that the formula is Used as a reference zero point;

[0099] Future No. sky( The increase in temperature influence for:

[0100] ;

[0101] in, For the first The future of the region The temperature of the day affects the increment, unit: m 3 , For the future Forecast daily average temperature for the day, unit: °C. Historical baseline temperature, unit: °C Temperature influence coefficient, unit: m 3 / ℃, This is the baseline correction amount, in meters. 3 ,Depend on The result is obtained through conversion.

[0102] It should be noted that temperature is a key meteorological factor affecting irrigation water demand. When temperatures rise, vegetation transpiration and soil evaporation increase, leading to a rise in water demand. This step establishes a linear regression model between temperature and water demand. This quantitatively characterizes the marginal impact of temperature changes on water demand. To predict the degree to which temperature deviates from historical baseline temperatures, This represents the increase in water demand caused by a 1°C increase. This is the correction term for the regression intercept after shifting it to the historical baseline temperature. and All data are derived from regression analysis of historical temperature and water consumption data, and do not rely on forecast data; forecast temperature It is only used to substitute into the regression equation to calculate the future temperature impact increment.

[0103] It should be further explained that, Take the arithmetic mean of the historical daily average temperatures over the 30 consecutive days preceding the current moment; and Linear regression of historical daily average temperature and historical daily irrigation water consumption over the preceding 30 days using the least squares method yielded the results. The 30-day window length strikes a balance between sensitivity to temperature change trends and regression stability, and the sample size meets the basic statistical requirements for linear regression. The standard least squares formula was used.

[0104] ;

[0105] in and These are the arithmetic averages of historical temperature and historical water consumption over the past 30 days, respectively. Converted .

[0106] In one embodiment of the present invention: the forecast daily average temperature for the first day (August 2nd). Historical benchmark temperature The average daily temperature for the 30 consecutive days prior to August 1st (July 3rd to August 1st) was taken as 25℃. Least squares regression (30 historical sample points, coefficient of determination) was used. )have to m 3 / ℃, m 3 ,Depend on m 3 . m 3 .

[0107] Step S2.3: Combine the baseline irrigation water demand with the incremental effect of temperature to calculate the predicted irrigation water demand.

[0108] Based on the baseline irrigation demand and the incremental effect of temperature, the predicted irrigation demand for each irrigation area for the next 7 days is calculated using the following method:

[0109] ;

[0110] in For the first The future of the region Predicted irrigation water demand per day, unit: m³ 3 , Ensure that the predicted value is non-negative. For the first The future of the region The temperature of the day affects the increment, unit: m 3 , For the first The irrigation area The standard irrigation water requirement per day, in cubic meters. 3 .

[0111] This step integrates the baseline irrigation water demand with the incremental impact of temperature, enabling a comprehensive depiction of the changing patterns of irrigation water demand and providing a foundational level. Adjusted from the perspective of temperature. Cut-off to avoid low-temperature scenarios Negative values ​​produce non-physical results.

[0112] In one embodiment of the present invention: On the first day in the future (August 2nd), in region A: m 3 Similarly, for region B, on day 1 in the future... m 3 Region C, Day 1 of the Future m 3 Repeating the above calculations yields the results for each region for each of the next 7 days. .

[0113] Step S3: Calculate the daily available water volume and establish equations for single-period water source constraints, multi-period cumulative water source constraints, and dynamic changes in water storage.

[0114] It should be noted that the above steps have outputted the predicted irrigation water demand, but this is an ideal demand. In actual scheduling, the available water supply for irrigation is limited, and the available water volume in each period is constrained by the initial water storage. Furthermore, excessive water withdrawal in the early stages will directly reduce the available water volume in the later stages. Therefore, this step ensures that the scheduling plan remains feasible throughout the entire optimization cycle based on real-time water storage, replenishment rate, and forecasted precipitation.

[0115] S3.1: Calculate the daily available water volume based on real-time water storage, replenishment rate, and precipitation replenishment.

[0116] Based on the current real-time water storage, replenishment rate, and precipitation forecast, the daily available water volume after the current moment is calculated recursively, as follows:

[0117] ;

[0118] in, For the first Daily available water volume For the first Daily water storage capacity, unit: m³ 3 , Replenishment rate, unit: m 3 / sky; The time step is in days, and we take 1 day. For the first Rainfall replenishment, unit: m 3 It is calculated from the precipitation forecast: , The water source catchment area, in m². The runoff coefficient, Rainfall amount, in mm.

[0119] It should be noted that the available water volume in the current period is equal to the initial water storage volume. and current supply It consists of two parts.

[0120] S3.2: Establish single-period water source constraints based on daily available water volume.

[0121] A single-period water source constraint is established based on the available water volume of the day, stipulating that the total water supply of the day shall not exceed the available water volume of the day.

[0122] ;

[0123] in For the first The first region Daily water supply volume, unit: m³ 3 , are decision variables in the optimization model; This represents the total number of irrigated areas. For the first Daily available water volume.

[0124] It should be noted that the total water supply of all irrigated areas on a given day must not exceed the available water volume for that day, ensuring that the daily water supply plan is within the water source capacity for that day. It refers to the first in the optimization model The irrigation area The required water supply volume for each day, in cubic meters (m³). 3 , is the decision variable of the optimization model, that is, the unknown amount of water that should be supplied to each region every day, which is ultimately calculated by the model through the genetic algorithm.

[0125] Step S3.3: Establish multi-time period cumulative water source constraints.

[0126] Based on the current water storage and replenishment rate, establish a multi-period cumulative water source constraint, which stipulates that the total cumulative water supply from the current time to the end of the optimization date must not exceed the sum of the current water storage and cumulative replenishment.

[0127] ;

[0128] in, To optimize the total number of time periods, , This represents the total number of irrigated areas. For the first The first region The decision on water supply for the day The first day's water storage volume is recorded as the current water storage volume. For the replenishment rate, For the first Rainfall replenishment The time step is in days, and we take 1 day.

[0129] From the current moment to the [time period] The total cumulative water supply for the day shall not exceed the current water storage capacity. The sum of the cumulative replenishment over all days. Single-period water source constraints ensure that the daily water supply does not exceed the available water volume for that day. However, this constraint cannot reflect the impact of excessive water withdrawal in the early stages on the available water volume in the later stages. Even if the single-period constraints for each day are met, scenarios may still occur where excessive water withdrawal in the early stages leads to insufficient water storage in the later stages, or where subsequent periods, although not exceeding the supply limit, cannot meet the minimum water supply demand due to insufficient water storage. Therefore, it is necessary to introduce multi-period cumulative water source constraints to limit the total cumulative water supply from a global perspective throughout the entire optimization cycle, ensuring that it does not exceed the sum of the current water storage and the total cumulative replenishment over the entire cycle, thereby guaranteeing the feasibility of the scheduling scheme across all time scales.

[0130] Step S3.4: Establish the equation for the dynamic change of water storage.

[0131] Establish a dynamic change equation for water storage to describe the dynamic evolution of water storage from day to day.

[0132] ;

[0133] in, Indicates the first Daily water storage capacity, unit: m³ 3 , Replenishment rate, unit: m 3 / sky; The time step is in days, and we take 1 day. For the first Rainfall replenishment, unit: m 3 It is calculated from the precipitation forecast. This represents the total number of irrigated areas. For the first The first region The decision on water supply for the day To optimize the total number of time periods, , For the first Natural water storage capacity.

[0134] It should be noted that the initial water storage volume on the following day Equal to the initial water storage volume of the day Add the daily replenishment amount and subtract the total daily water supply. This equation and formula... and as well as This constitutes a complete dynamic constraint system for water sources. (Formula) Define the available water volume for the day, using the formula. Constrain the daily water supply, formula Constraint on cumulative water supply, formula Drive the evolution of water storage status. Through formula It can be obtained recursively .

[0135] In one embodiment of the present invention: the current water storage capacity of the reservoir m 3 The actual level gauge reading shows the replenishment rate. m 3 / day, actual flow meter measurement, rainfall replenishment amount m 3 / day, the forecast precipitation for the day is 0. sky, , .

[0136] Step S3.1: Initial water storage on day 1 m 3 Available water volume for the day m 3 .

[0137] Step S3.2: Single-period water source constraint m 3 .

[0138] Step S3.3: 7-day cumulative water source constraint m 3 .

[0139] Step S3.4: If the total water supply on day 1 is 164m³ 3 The initial water storage volume on the second day m 3 .

[0140] Step S4: Calculate the irrigation demand deficit, water waste, and water supply cost. Determine the weighting coefficients using the analytic hierarchy process (AHP) and construct a benchmarked multi-objective optimization scheduling model.

[0141] It should be noted that step S2 provides a forecast of irrigation water demand. Step S3 provides single-period and cumulative constraints for water resources. To evaluate the effectiveness of the scheduling scheme, it needs to be measured from three dimensions: demand satisfaction, water resource utilization efficiency, and water supply economics. These three sub-objectives have different dimensions; the deficit and waste are represented by m. 3 The water supply cost is in yuan and cannot be directly added together. This step first uses the water supply volume of each irrigation area at each time period as the decision variable to be optimized. Based on the predicted irrigation water demand obtained in step S2 and the decision variable, the deficit and waste are calculated. Based on the decision variable and the unit water supply cost, the water supply cost is calculated. Then, after benchmarking and dimensionless processing of each sub-objective, a weighted sum is calculated to construct a multi-objective optimization scheduling model.

[0142] Step S4.1: Calculate the irrigation demand deficit.

[0143] The irrigation demand deficit is calculated based on the predicted irrigation water demand and the decided water supply.

[0144] It should be noted that the irrigation demand deficit... When deciding on water supply Less than the predicted irrigation water demand At that time, a deficit occurs. The sum of deficits across all regions and time periods, in meters. 3 . The smaller the value, the higher the degree of demand satisfaction.

[0145] ;

[0146] In the formula, This represents the irrigation deficit, expressed in meters. 3 , For the first The first region Predicted irrigation water demand per day, in cubic meters. 3 ; For the first The first region Daily water supply volume, in cubic meters 3 , This represents the total number of irrigated areas. To optimize the total number of time periods, .

[0147] It should be noted that a deficit occurs and accumulates when the water supply is less than the water demand; maximizing the irrigation demand satisfaction rate is equivalent to the amount of the deficit. minimize.

[0148] Step S4.2: Calculate the amount of water wasted.

[0149] The amount of water wasted is calculated based on the predicted irrigation water demand and the decided water supply.

[0150] It should be noted that the amount of water wasted When deciding on water supply Greater than the predicted irrigation water demand This results in waste. The sum of waste across all regions and time periods, in meters. 3 . The smaller the value, the higher the water resource utilization efficiency.

[0151] ;

[0152] In the formula, Water wastage, expressed in cubic meters (m³)3 , For the first The first region Daily water supply volume, in cubic meters 3 , For the first The first region Predicted irrigation water demand per day, in cubic meters. 3 , This represents the total number of irrigated areas. To optimize the total number of time periods, .

[0153] This indicates that waste occurs and accumulates when the water supply exceeds the water demand; the waste amount is 0 when the water supply is less than or equal to the water demand. and Positive values ​​will not be generated simultaneously in the same time period and region. Optimization requires finding a balance between "undersupply" and "oversupply".

[0154] Step S4.3: Calculate the water supply cost.

[0155] Based on the decision on water supply and unit water supply cost Calculate water supply costs .

[0156] It should be noted that the water supply cost The total cost of water supply is determined by the water supply volume for each region and time period. Its unit water supply cost The sum of the products, in units of yuan. The smaller the value, the better the economic efficiency of water supply.

[0157] ;

[0158] In the formula, For water supply costs, For the first The first region Daily unit water supply cost, unit: yuan / m³ 3 It is calculated by adding energy consumption costs to the local water fee standard. For the first The first region Daily water supply volume, in cubic meters 3 , This represents the total number of irrigated areas. To optimize the total number of time periods, .

[0159] It should be noted that the total water supply cost is the sum of the products of the water supply volume decided for each region and time period and the unit water supply cost.

[0160] Step S4.4: Determine the weight coefficients using the analytic hierarchy process (AHP), perform dimensionless benchmarking on the three sub-objectives, and construct the benchmarked objective function.

[0161] The irrigation demand deficit, water waste, and water supply cost are benchmarked and dimensionless, and a weighted summation is used to construct the objective function.

[0162] It should be noted that the three sub-objectives have different dimensions ( , For m 3 , Since these are elements, they cannot be directly added. After converting each sub-objective to a dimensionless value using a benchmarking factor, a weighted sum is calculated:

[0163] ;

[0164] in The overall target value is dimensionless. These are the first weighting coefficient, the second weighting coefficient, and the third weighting coefficient, respectively. They are dimensionless and satisfy the following conditions: , For the irrigation demand deficit, For the amount of water wasted, For water supply costs, This serves as a baseline value for irrigation demand deficit. This serves as a benchmark for water waste. This is the benchmark value for water supply costs; when , or When the value is 0, the corresponding baseline value is 1;

[0165] Methods for determining benchmark values:

[0166] The maximum water shortage when there is no water supply at all;

[0167] : For the first The maximum daily historical irrigation water consumption for each region, in cubic meters. 3 ; ;

[0168] ;

[0169] It should be noted that, in this invention, the reference value , , Its function is to convert three sub-objectives with different dimensions into dimensionless scalars, enabling them to be weighted and summed. The selection of the benchmark value follows the extreme scenario method, that is, selecting the maximum value of each sub-objective under physical limits as the benchmark;

[0170] :Pick Its physical meaning is: if there is no water supply at all, that is, all... The deficit in each time period The sum of these values ​​represents the total predicted irrigation water demand for all regions and all time periods. This is the maximum possible deficit, and the actual deficit can be calculated based on this. Mapped to Interval, dimensionless value The larger the value, the more severe the deficiency.

[0171] :Pick ,in For the first The maximum daily historical irrigation water consumption for each region, calculated from the 90 consecutive days preceding the current moment. The maximum value is obtained by summing up the data daily. Its physical meaning is: if water is supplied daily at the historical maximum supply (i.e.,...) ),and Much larger If actual needs are not considered at all, the amount of waste per day is approximately... The sum of these values ​​represents the maximum daily historical irrigation water consumption for all regions and time periods. This is the upper limit of potential waste, and the actual amount of waste can be calculated based on this. Mapped to Interval.

[0172] :Pick Its physical meaning and Corresponding to: the water supply cost incurred when supplying water based on the historical maximum water supply volume. The actual water supply cost can be calculated based on this benchmark. Mapped to Interval.

[0173] The essence of benchmarking is to provide a unified metric for each sub-objective. Choosing the maximum value under extreme scenarios as the benchmark ensures that the actual sub-objective value divided by the benchmark value always falls within the range of the benchmark value. The interval ensures the comparability of dimensionless values ​​for each sub-objective, preventing the optimization direction from being dominated by the excessively large magnitude of a single sub-objective. Simultaneously, the physical meaning of extreme scenarios is clear, allowing those skilled in the art to directly calculate based on actual data.

[0174] The Analytic Hierarchy Process (AHP) is used to determine this, and the specific steps are as follows:

[0175] Several experts in the field of water resource allocation were invited to discuss the three sub-objectives (irrigation demand deficit). Water waste Water supply costs The importance of each factor is compared pairwise. A 1-9 scale is used: 1 indicates equal importance, 3 indicates slightly important, 5 indicates significantly important, 7 indicates strongly important, and 9 indicates extremely important. For example, if experts believe that "meeting needs is significantly more important than cost control," then the comparison value is 5.

[0176] Construct a judgment matrix by taking the geometric mean of the pairwise comparison results of all experts. ,in Indicates the first The sub-goals are relative to the first The importance of individual goals.

[0177] Calculate the judgment matrix Maximum eigenvalue The corresponding feature vector, after normalization, becomes the weight vector. .

[0178] Perform a consistency check: Calculate the consistency ratio. ,in , As a random consistency index, a 3D matrix .when If the weight judgment passes the consistency test, it is considered to have passed the consistency test; otherwise, experts need to be consulted to readjust the judgment matrix.

[0179] Step S4.5: Supplement the non-negative water supply constraints and minimum water supply constraints to establish complete constraint conditions.

[0180] Based on the single-period water source constraints and cumulative water source constraints in step S3, establish complete water source constraints for the multi-objective optimization model, including non-negative water supply constraints and minimum water supply constraints.

[0181] It should be noted that step S3 has established single-period water source constraints and multi-period cumulative water source constraints, but these two types of constraints only restrict the water balance relationship between water supply and available water. In the multi-objective optimization model, additional decision variables are needed. The basic physical constraints that it must meet are: first, the water supply cannot be negative; and second, each region must have a minimum irrigation demand. Therefore, this step combines the single-period water source constraints and cumulative water source constraints from step S3 with the non-negative water supply constraints and minimum water supply constraints to form the complete constraints of the multi-objective optimization model.

[0182] The constraints are as follows:

[0183] ;

[0184] ;

[0185] ;

[0186] ;

[0187] in For the first Minimum water supply capacity for each region, unit: m³ 3 Take 70% to 80% of the lowest daily historical irrigation water consumption over 90 consecutive days; Replenishment rate, unit: m 3 / sky, For the first Rainfall replenishment The time step is in days, and we take 1 day. For the first The first region Daily water supply volume, unit: m³ 3 , are decision variables in the optimization model; This represents the total number of irrigated areas. For the first The first region The decision on water supply for the day For the first Daily available water volume The first day's water storage is recorded as the current water storage.

[0188] It should be noted that, Due to single-period water source constraints, To address the cumulative water source constraints over multiple time periods, Ensure a non-negative water supply. To ensure minimum irrigation needs in each region, the lowest daily water consumption value over the previous 90 days reflects the lowest level of water demand in that region during the same historical period. This minimum value typically occurs during periods of low temperature, high soil moisture, or slow vegetation growth, representing the basic water consumption required to maintain vegetation survival. A margin of 70%–80% is added to account for the fact that in actual scheduling, it is impossible to precisely supply water according to the historical minimum level, due to minimum flow limitations in the pipeline network, operational deviations in valve opening and closing, and water allocation losses between different regions. Reserving a 20%–30% margin ensures that even with these deviations, the actual water supply will not fall below the minimum threshold required for vegetation survival, preventing vegetation from withering due to insufficient water supply.

[0189] In one embodiment of the present invention, urban park greening irrigation includes three irrigation areas: a lawn area ( Garden area ( ), forest area ( The current time is 00:00 on August 1, 2025. (Optimized time period) Day; Predicted irrigation water demand: For each region on the first day of the future, the following figures apply. m 3 , m 3 , m 3 The sum of the predicted water demand for each region over the next 7 days is: m 3 , m 3 , m 3 ,total m 3 In this embodiment, to simplify the calculation, it is assumed that the weather forecast data for 7 days is the same, so the predicted water demand for each day is the same.

[0190] Unit water supply cost in each region: Yuan / m 3 , Yuan / m 3 , Yuan / m 3 .

[0191] Maximum daily historical irrigation water consumption in each region: m 3 , m 3 , m 3 (Determined by the maximum daily irrigation water consumption in the previous 90 days).

[0192] Weighting coefficients: determined using the analytic hierarchy process (AHP). , , (Conformity ratio) (Passed the consistency test).

[0193] Water resource constraints: Available water volume in a single time period and the 7-day cumulative water source constraint limit m 3 .

[0194] Minimum water supply for each region: m 3 , m 3 , m 3 (Take 70% to 80% of the lowest daily irrigation water consumption in the previous 90 days).

[0195] Baseline value calculation: ; ; ;

[0196] Multi-objective optimization scheduling model:

[0197] Objective function: ;

[0198] in: ; ; ;

[0199] Constraints: ; ; ; .

[0200] Step S5: Under the complete water source constraint, use a genetic algorithm to solve the multi-objective optimization scheduling model to obtain the decision water supply quantity and water supply priority.

[0201] It should be noted that step S4 has constructed a complete optimization mathematical model, but the decision variables in the model... Specific values ​​have not yet been determined. This step uses a genetic algorithm to numerically solve the model, obtaining the decision water supply volume and water supply priority for each region.

[0202] A genetic algorithm is used to solve the multi-objective optimization scheduling model, with the decision variables having the following dimensions. The population size is 5 to 10 times the dimension of the variable, the crossover probability is 0.7 to 0.9, the mutation probability is 0.05 to 0.15, and the maximum number of iterations is 800 to 1500 generations; This represents the total number of irrigated areas. To optimize the total number of time periods, ;

[0203] Output after solving by the genetic algorithm:

[0204] Decision on water supply Unit: m 3 ;

[0205] Water supply priority: determined based on the urgency of the predicted irrigation water demand in each region, with higher priority given to regions with greater water demand.

[0206] It should be noted that the above parameter values ​​are all commonly used ranges based on engineering experience in this field, and those skilled in the art can adjust them within these ranges according to the actual problem scale. The objective function of the benchmarked multi-objective optimization scheduling model is used as the fitness function; a smaller fitness value indicates a better individual. For individuals that do not meet the single-period water source constraint, cumulative water source constraint, non-negative water supply constraint, or minimum water supply constraint, a constraint processing method known in the art is used to increase their fitness value, thereby eliminating them in the selection operation. When the maximum number of iterations is reached or the rate of change of the optimal fitness value is less than 0.1% for 50 consecutive generations, the algorithm is considered to have converged and the iteration terminates. After the algorithm terminates, the individual with the smallest fitness value in the final population is taken as the optimal solution, and its corresponding decision water supply is the optimal water supply for each irrigation area at each time period.

[0207] In one embodiment of the present invention: population size: 150; crossover probability: 0.8; mutation probability: 0.1; maximum number of iterations: 1000 generations;

[0208] Through iterative optimization, the decision-making water supply for each region was converged. m 3 , m 3 , m 3 Total water supply m 3 m 3 (Single time period constraint), 7-day cumulative water supply m 3 m 3 (Cumulative constraint). Based on the predicted irrigation water demand for each region. The urgency of water supply determines the priority of water supply: Area A ( m 3 > Region C ( m 3 > Region B ( m 3 In this embodiment, for the sake of simplifying the calculation, it is assumed that the water supply is the same for 7 days.

[0209] Step S6: Generate and output a scheduling decision scheme based on the decision water supply volume and water supply priority combined with the hourly water consumption coefficient, and perform rolling optimization when the preset update cycle is reached.

[0210] It should be noted that the decision water supply output in step S5 This involves daily water supply data for each region, which needs to be converted into executable scheduling commands. Hourly water consumption coefficient. The water consumption data, obtained in step S2.1 from hourly historical water consumption statistics, reflects the distribution of water demand in different time periods. This step generates a water supply schedule by allocating the hourly water supply based on the decision water supply volume and hourly water consumption coefficient, and generates a complete dispatching decision scheme by combining water supply priority and dynamic change equations of water storage.

[0211] The specific method is as follows:

[0212] The first in each region Daily water supply decision Distributed to each hour:

[0213] ;

[0214] in For the first The first region Heavenly Hourly water supply allocation, unit: m 3 , For the first The first region The hourly water consumption coefficient is obtained from hourly historical irrigation water consumption statistics over 90 consecutive days and is dimensionless. As the normalization factor, For the first The first region Daily water supply volume, unit: m³ 3 .

[0215] The scheduling decision-making scheme includes: water supply schedule: water intake The longest continuous period is used as the recommended water supply window, listing the recommended water supply periods for each irrigation area; Water supply recommendations: Listing the recommended water supply volume for each irrigation area for each time period. Water supply priority: output from step S5; Irrigation water source status early warning: based on the dynamic change equation of water storage. Calculated To determine whether the water storage capacity of the water source is below the safety threshold;

[0216] A preset update cycle is set. When the preset update cycle is reached, the process returns to step S1 to update historical irrigation water consumption, historical temperature data, weather forecast data, real-time water storage and replenishment rate, and steps S2 to S6 are executed again.

[0217] It should be noted that the safety threshold is determined as follows: 1.2 times the sum of the minimum guaranteed water supply for each region over the next 7 days is taken as the safety threshold. This means that when the water storage is lower than this threshold, in the event of a sudden situation, such as continuous drought or pipeline failure, the minimum irrigation needs of each region cannot be guaranteed. Those skilled in the art can adjust the coefficient appropriately based on the actual water source regulation capacity. The 1.2 times coefficient is derived from engineering practice experience, representing a 20% safety margin reserved on top of the minimum water supply demand to cope with abnormal situations such as sudden drought or pipeline leakage. This coefficient can be adjusted within the range of 1.0 to 1.5 according to local climate conditions and pipeline reliability.

[0218] It should be noted that the selection of the update cycle is based on the update frequency of meteorological forecast data. Currently, meteorological forecasts are typically updated every 6 to 24 hours. If the update cycle is shorter than 6 hours, the meteorological forecast data has not yet been updated, and re-optimization cannot obtain new and effective information. If the update cycle is too long, it cannot respond promptly to dynamic fluctuations in weather changes and water demand. Using one day ensures that the meteorological forecast data has been updated for each optimization, while allowing sufficient optimization calculation time. Furthermore, in engineering practice, irrigation scheduling is usually based on a daily decision unit, making it most convenient for managers to execute water supply plans on a daily basis. Therefore, those skilled in the art can appropriately adjust the one-day baseline based on the actual update frequency of meteorological forecasts and the rate of change in water demand in each irrigation area. When hourly forecasts are used and computational resources are sufficient, the interval can be shortened to 12 hours or 6 hours; when water demand changes slowly, it can be extended to 2-3 days, but one day is preferred.

[0219] In one embodiment of the present invention: the peak water supply is between 10:00 and 16:00, and the hourly water supply is allocated by a formula. The scheduling scheme is as follows: Area A supplies 80m³ of water daily from 10:00 to 16:00. 3 Area B receives 40m³ of water daily from 8:00 to 10:00. 3 Area C receives 44m³ of water daily from 17:00 to 19:00. 3 . m 3 144 m above the safety threshold 3 Status is normal, safety threshold: .

[0220] See Figure 2 This invention discloses an irrigation water resource scheduling system based on data prediction, comprising:

[0221] The data acquisition module is used to obtain hourly historical irrigation water consumption and historical temperature data for each irrigation area for 90 consecutive days before the current moment, obtain meteorological forecast data for 7 consecutive days after the current moment, and obtain real-time water storage and replenishment rate of irrigation water sources.

[0222] The water demand forecasting module is used to obtain the daily historical irrigation water volume by summarizing the hourly historical irrigation water volume on a daily basis, and to recursively calculate the baseline irrigation water demand for each irrigation area using the exponential smoothing method; to establish a temperature impact correction model based on the historical temperature data, and to calculate the temperature impact increment by combining the forecast temperature in the meteorological forecast data; and to superimpose the baseline irrigation water demand with the temperature impact increment to obtain the predicted irrigation water demand for each irrigation area.

[0223] The water source constraint module is used to calculate the daily available water volume based on the real-time water storage, replenishment rate and forecast precipitation in the meteorological forecast data, establish a single-period water source constraint based on the daily available water volume, establish a multi-period cumulative water source constraint based on the current water storage and cumulative replenishment, and establish a dynamic change equation of water storage describing the daily change law of water source water storage.

[0224] The model building module is used to calculate the irrigation demand deficit and water waste based on the predicted irrigation water demand and the decision variables, taking the water supply volume of each irrigation area at each time period as the decision variables to be optimized; calculate the water supply cost based on the decision variables and the unit water supply cost; and construct a benchmarked multi-objective optimization scheduling model by performing benchmarked dimensionless processing on the irrigation demand deficit, water waste and water supply cost respectively.

[0225] The optimization solution module is used to solve the benchmarked multi-objective optimization scheduling model using a genetic algorithm, taking the single-time water source constraint, multi-time cumulative water source constraint, water storage dynamic change equation, water supply non-negativity constraint, and minimum water supply constraint as complete constraints, to obtain the decision water supply and water supply priority for each irrigation area.

[0226] The scheme output module is used to allocate the daily decision water supply to each hour based on the decision water supply volume and water supply priority, combined with the hourly water consumption coefficient obtained from the hourly historical irrigation water consumption statistics, to generate a water supply schedule and scheduling decision scheme and output it; when the preset update cycle is reached, it returns to the data acquisition module for rolling optimization.

[0227] This invention achieves forward-looking and accurate prediction of irrigation water demand. It utilizes historical irrigation water consumption and historical temperature data, employing exponential smoothing and least squares linear regression to establish a baseline irrigation water demand recursive model and a temperature-corrected model, respectively. Combined with forecasted temperatures from meteorological data, the invention calculates the predicted irrigation water demand for each irrigation area over the next 7 days. The exponential smoothing coefficient is automatically optimized using a grid search method within the range of 0.2 to 0.4, aiming to minimize the fitting error of historical data. The temperature regression model is dynamically updated based on rolling 30-day historical data, enabling the prediction model to adaptively track changes in water consumption patterns in each irrigation area. Unlike existing passive scheduling methods that rely on fixed quotas or threshold triggers, this invention achieves forward-looking quantitative prediction of future water demand, transforming scheduling decisions from reactive post-event responses to proactive pre-event planning.

[0228] A complete dynamic water source constraint system was constructed to ensure the feasibility of the scheduling scheme. Based on real-time water storage, replenishment rate, and forecasted precipitation, this invention establishes a triple constraint mechanism: single-period water source constraint, multi-period cumulative water source constraint, and a dynamic water storage change equation. The single-period water source constraint ensures that the daily water supply does not exceed the available water volume for that day. The multi-period cumulative water source constraint ensures that the cumulative water supply over the entire optimization cycle does not exceed the sum of the current water storage and the cumulative replenishment. The dynamic water storage change equation describes the daily evolution of water storage. By organically combining these three constraints, the scheduling scheme is ensured to always remain within the water supply capacity range from three dimensions: daily water volume limit, full-cycle cumulative water use limit, and dynamic water volume evolution. This effectively avoids the problem of water source depletion or water supply interruption due to excessive water use in the early stages. Simultaneously, this invention adds a minimum water supply constraint for each irrigation area, taking 70%–80% of the historical lowest water use as the minimum guaranteed water supply, forcibly retaining the basic irrigation water required for vegetation survival, while also considering the ecological safety baseline.

[0229] This invention achieves multi-objective comprehensive trade-offs and global optimal scheduling. It simultaneously incorporates irrigation demand deficit, water waste, and water supply cost into the optimization objectives. Employing an extreme scenario benchmarking dimensionless processing method, it uses the maximum deficit under no water supply, the historical maximum water supply, and the corresponding maximum water supply cost as benchmark values. This unifies the three sub-objectives with different dimensions into dimensionless scalars, solving the problem in existing technologies where demand satisfaction rate, water resource utilization rate, and water supply economics cannot be directly compared and comprehensively balanced. Based on this, the analytic hierarchy process (AHP) is used to determine the weight coefficients of each sub-objective. Through pairwise comparisons of the judgment matrix by experts and consistency checks, the weight allocation has a theoretical basis and interpretability, avoiding the subjectivity of arbitrary weight assignment. This achieves globally optimal scheduling of three conflicting objectives within the same framework.

[0230] This invention generates a finely detailed hourly scheduling scheme that can be directly implemented and establishes a rolling optimization mechanism. Based on the daily water supply decision obtained from a genetic algorithm and the water supply priority of each region, combined with the hourly water consumption coefficient obtained from hourly historical irrigation water consumption statistics, the daily water supply is allocated to each hour. This generates a complete scheduling decision scheme that includes a water supply schedule, time-segmented water supply suggestions, water supply priorities, and water source status early warning information, allowing the optimization results to be directly converted into executable operational instructions. Simultaneously, this invention sets a preset update cycle. When the cycle arrives, the latest data is automatically collected and the entire optimization process is re-executed. This enables the scheduling scheme to be dynamically adjusted based on weather forecast updates and actual water consumption feedback, maintaining the continuous optimality of the decision and improving the system's engineering practicality and environmental adaptability.

[0231] The above are merely preferred embodiments of the present invention and are not intended to limit the present invention. Various modifications and variations can be made to the present invention by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A data-based method for irrigation water resource scheduling, characterized in that, include: Step S1: Obtain the hourly historical irrigation water consumption and historical temperature data for each irrigation area for 90 consecutive days before the current moment, obtain the weather forecast data for 7 consecutive days after the current moment, and obtain the real-time water storage and replenishment rate of the irrigation water source. Step S2: Based on the hourly historical irrigation water consumption, the daily historical irrigation water consumption is obtained by summarizing the daily data. The baseline irrigation water demand for each irrigation area is calculated recursively using the exponential smoothing method. A temperature influence correction model is established based on the historical temperature data, and the temperature influence increment is calculated by combining the forecast temperature in the meteorological forecast data. The baseline irrigation water demand and the temperature influence increment are superimposed to obtain the predicted irrigation water demand for each irrigation area. Step S3: Based on the real-time water storage, replenishment rate and forecast precipitation in the meteorological forecast data, calculate the daily available water volume, establish a single-period water source constraint based on the daily available water volume, establish a multi-period cumulative water source constraint based on the current water storage and cumulative replenishment, and establish a dynamic change equation for water storage that describes the daily change pattern of water source water storage. Step S4: Using the water supply volume of each irrigation area at each time period as the decision variable to be optimized, calculate the irrigation demand deficit and water waste based on the predicted irrigation water demand and the decision variable; calculate the water supply cost based on the decision variable and the unit water supply cost. After standardizing and dimensionlessly processing the irrigation demand deficit, water waste, and water supply cost, a weighted sum is calculated to construct a standard multi-objective optimization scheduling model. Step S5: Using the single-period water source constraint, multi-period cumulative water source constraint, water storage dynamic change equation, water supply non-negativity constraint, and minimum water supply constraint as complete constraints, the genetic algorithm is used to solve the benchmark multi-objective optimization scheduling model to obtain the decision water supply and water supply priority for each irrigation area. Step S6: Based on the decision water supply volume and water supply priority, and combined with the hourly water consumption coefficient obtained from the hourly historical irrigation water consumption statistics, allocate the daily decision water supply volume to each hour to generate a water supply schedule and scheduling decision scheme and output it; when the preset update cycle is reached, return to step S1 for rolling optimization.

2. The irrigation water resource scheduling method based on data prediction according to claim 1, characterized in that, The hourly historical irrigation water consumption refers to the hourly irrigation water consumption of each irrigation area for the 90 consecutive days prior to the current moment, collected hourly by smart water meters installed on the water supply pipelines of each irrigation area; the historical temperature data refers to the daily average temperature for the 90 consecutive days prior to the current moment; the meteorological forecast data includes the forecast daily maximum temperature, forecast daily minimum temperature, and forecast precipitation for the 7 consecutive days following the current moment; the real-time water storage is read by a level sensor installed in the irrigation water source, and the replenishment rate is read by a flow meter installed on the replenishment pipeline.

3. The irrigation water resource scheduling method based on data prediction according to claim 2, characterized in that, The recursive calculation method for the baseline irrigation water requirement is as follows: The hourly historical irrigation water consumption is summarized on a daily basis to obtain the daily historical irrigation water consumption; ； in For the first The irrigation area The standard irrigation water requirement per day, in cubic meters. 3 , For the first The irrigation area Daily historical irrigation water consumption per day, in cubic meters. 3 , The smoothing coefficient is dimensionless. For the first The irrigation area The baseline irrigation water requirement for the day; when hour, The average daily irrigation water usage over the seven consecutive days preceding the current moment is taken. Seven days corresponds to a complete week, which covers the difference in water usage between weekdays and weekends. The value ranges from 0.2 to 0.4, and the values ​​are iterated in increments of 0.

01. The optimal value is determined by minimizing the fitting error of the historical data for the 90 consecutive days prior to the current moment.

4. The irrigation water resource scheduling method based on data prediction according to claim 3, characterized in that, The method for establishing the temperature impact correction model is as follows: For the A least-squares linear regression was performed on the historical daily average temperature of the 30 consecutive days preceding the current time in an irrigation area and the historical daily irrigation water consumption of that area to obtain the temperature influence coefficient and the linear regression intercept; the historical baseline temperature was calculated as the arithmetic mean of the historical daily average temperatures of the 30 consecutive days; the baseline correction was obtained by converting the regression intercept and the temperature influence coefficient; The future of the region The method for calculating the incremental impact of daytime temperature is as follows: ； in, For the first The future of the region The temperature of the day affects the increment, unit: m 3 , For the future Forecast daily average temperature for the day, unit: °C. Historical baseline temperature, unit: °C Temperature influence coefficient, unit: m 3 / ℃, This is the baseline correction amount, in meters. 3 ,Depend on The conversion yields, This is the intercept for linear regression.

5. The irrigation water resource scheduling method based on data prediction according to claim 4, characterized in that, The method for calculating the predicted irrigation water demand is as follows: ； in For the first The future of the region Predicted irrigation water demand per day, unit: m³ 3 , Ensure that the predicted value is non-negative. For the first The future of the region The temperature of the day affects the increment, unit: m 3 , For the first The irrigation area The standard irrigation water requirement per day, in cubic meters. 3 .

6. The irrigation water resource scheduling method based on data prediction according to claim 1, characterized in that, The calculation of the daily available water volume is as follows: ； in, For the first Daily available water volume For the first Daily water storage capacity, unit: m³ 3 , Replenishment rate, unit: m 3 / sky; The time step is in days, and we take 1 day. For the first Rainfall replenishment, unit: m 3 It is calculated by combining the forecast precipitation with the catchment area and runoff coefficient; The single-time water source constraint is: ； in, For the first The first region Daily water supply volume, unit: m³ 3 , are decision variables in the optimization model; This represents the total number of irrigated areas. The multi-time period cumulative water source constraint is: ； in, The water volume on the first day is recorded as the water volume at the current moment. The equation for the dynamic change of water storage is: ； in, Indicates the first Daily water storage capacity, unit: m³ 3 , To optimize the total number of time periods, .

7. The irrigation water resource scheduling method based on data prediction according to claim 6, characterized in that, The irrigation demand deficit is: ； The amount of water wasted is: ； The water supply cost is: ； The objective function of the benchmark multi-objective optimization scheduling model is: ； in, This represents the irrigation deficit, expressed in meters. 3 , For the first The first region Predicted irrigation water demand per day, in cubic meters. 3 ; For the first The first region Daily water supply volume, in cubic meters 3 , This represents the total number of irrigated areas. To optimize the total number of time periods, ; Water wastage, expressed in cubic meters (m³) 3 , For water supply costs, For the first The first region Daily unit water supply cost, unit: yuan / m³ 3 It is calculated by adding energy consumption costs to the local water fee standard; The overall target value is dimensionless. These are the first, second, and third weighting coefficients, which are dimensionless and determined using the analytic hierarchy process (AHP), satisfying the following conditions: , This serves as a baseline value for irrigation demand deficit. This serves as a benchmark for water waste. This is the benchmark value for water supply costs; when , or When the value is 0, the corresponding baseline value is taken as 1.

8. The irrigation water resource scheduling method based on data prediction according to claim 7, characterized in that, The method for determining the benchmark value is as follows: The baseline value for irrigation demand deficit is the sum of the predicted irrigation water demand for each region over seven consecutive days. The baseline value for water waste is the sum of the products of the maximum daily historical irrigation water consumption in each region over 90 consecutive days and 7. The benchmark value for water supply cost is the sum of the products of the maximum daily historical irrigation water consumption in each region over 90 consecutive days and the unit water supply cost.

9. The irrigation water resource scheduling method based on data prediction according to claim 8, characterized in that, The complete constraints also include a non-negative water supply constraint and a minimum water supply constraint. The minimum water supply is taken as 70% to 80% of the lowest daily historical irrigation water consumption over the previous 90 consecutive days. The parameters of the genetic algorithm are: population size is 5 to 10 times the dimension of the decision variables, crossover probability is 0.7 to 0.9, mutation probability is 0.05 to 0.15, and the maximum number of iterations is 800 to 1500 generations. The hourly water supply allocation formula is as follows: ； in For the first The first region Heavenly Hourly water supply allocation, unit: m 3 , For the first The first region The hourly water consumption coefficient is dimensionless and is obtained from the statistical data of historical irrigation water consumption over 90 consecutive hours. The preset update cycle is 1 day.

10. An irrigation water resource scheduling system based on data prediction, characterized in that, include: The data acquisition module is used to obtain hourly historical irrigation water consumption and historical temperature data for each irrigation area for 90 consecutive days before the current moment, obtain meteorological forecast data for 7 consecutive days after the current moment, and obtain real-time water storage and replenishment rate of irrigation water sources. The water demand forecasting module is used to obtain the daily historical irrigation water volume by summarizing the hourly historical irrigation water volume on a daily basis, and to recursively calculate the baseline irrigation water demand for each irrigation area using the exponential smoothing method; to establish a temperature impact correction model based on the historical temperature data, and to calculate the temperature impact increment by combining the forecast temperature in the meteorological forecast data; and to superimpose the baseline irrigation water demand with the temperature impact increment to obtain the predicted irrigation water demand for each irrigation area. The water source constraint module is used to calculate the daily available water volume based on the real-time water storage, replenishment rate and forecast precipitation in the meteorological forecast data, establish a single-period water source constraint based on the daily available water volume, establish a multi-period cumulative water source constraint based on the current water storage and cumulative replenishment, and establish a dynamic change equation of water storage describing the daily change law of water source water storage. The model building module is used to calculate the irrigation demand deficit and water waste based on the predicted irrigation water demand and the decision variables, using the water supply volume of each irrigation area at each time period as the decision variables to be optimized; and to calculate the water supply cost based on the decision variables and the unit water supply cost. After standardizing and dimensionlessly processing the irrigation demand deficit, water waste, and water supply cost, a weighted sum is calculated to construct a standard multi-objective optimization scheduling model. The optimization solution module is used to solve the benchmarked multi-objective optimization scheduling model using a genetic algorithm, taking the single-time water source constraint, multi-time cumulative water source constraint, water storage dynamic change equation, water supply non-negativity constraint, and minimum water supply constraint as complete constraints, to obtain the decision water supply and water supply priority for each irrigation area. The scheme output module is used to allocate the daily decision water supply to each hour based on the decision water supply volume and water supply priority, combined with the hourly water consumption coefficient obtained from the hourly historical irrigation water consumption statistics, to generate a water supply schedule and scheduling decision scheme and output it; when the preset update cycle is reached, it returns to the data acquisition module for rolling optimization.

Images